AMEGHINIANA (Rev. Asoc. Paleontol. Argent.) - 41 (4): 521-534. Buenos Aires, 30-12-2004 ISSN 0002-7014 New information on the Tapejaridae (Pterosauria, Pterodactyloidea) and discussion of the relationships of this clade Alexander Wilhelm Armin KELLNER1 Abstract. A phylogenetic analysis indicates that the Tapejaridae is a monophyletic group of pterodactyloid pterosaurs, diagnosed by the following synapomorphies: premaxillary sagittal crest that starts at the anterior tip of the premaxilla and extends posteriorly after the occipital region, large nasoantorbital fenestra that reaches over 45% of the length between premaxilla and squamosal, lacrimal process of the jugal thin, distinct small pear- shaped orbit with lower portion narrow, and broad tubercle at the ventroposterior margin of the coracoid. Several cranial and postcranial characters indicate that the Tapejaridae are well nested within the Tapejaroidea, in sister group relationship with the Azhdarchidae. A preliminary study of the ingroup relationships within the Tapejaridae shows that Tupuxuara is more closely related to Thalassodromeus relative to Tapejara. At present tape- jarid remains have been found in the following deposits: Crato and Romualdo members of the Santana Formation (Aptian-Albian), Araripe Basin, Brazil; Jiufotang Formation (Aptian), Jehol Group of western Liaoning, China; and in the redbeds (Cenomanian) of the Kem Kem region, Morocco. An incomplete skull found in the Javelina Formation (Maastrichtian), Texas also shows several tapejarid features and might be a member of this clade. Although information is still limited, the present distribution of the Tapejaridae indicates that this clade of pterosaurs was not exclusive of Gondwana, and was more widespread than previously known. Resumen. NUEVA INFORMACIÓN SOBRE LOS TAPEJARIDAE (PTEROSAURIA, PTERODACTYLOIDEA) Y DISCUSIÓN SOBRE LAS RELACIONES DE ESTE CLADO. Un análisis cladístico realizado indica que Tapejaridae es un grupo monofilético de los pterosaurios pterodactiloideos y está diagnosticado por las siguientes sinapomorfías: cresta sagital del pre- maxilar comienza en el borde anterior del premaxilar y se extiende posteriormente hasta la región occipital, gran fenestra nasoantorbital que alcanza el 45% de la extensión entre el premaxilar y escamosal, proceso lacrimal del yugal delgado, órbita pequeña en forma de pera con la porción inferior estrecha, y un amplio tubérculo en el margen ventro posterior del coracoide. Ciertos caracteres craneanos y postcraneanos indican que los Tapejaridae pueden agruparse dentro de los Tapejaroidea en una relación de grupo hermano con los Azhdarchidae. Un es- tudio preliminar de las relaciones del grupo interno dentro de los Tapejaridae muestra que Tupuxuara está más cercanamente relacionado a Thalassodromeus que a Tapejara. Hasta el presente los restos de tapejáridos han sido hallados en los siguientes depósitos: en los miembros Crato y Romualdo de la Formación Santana (Aptiano- Albiano), Cuenca de Araripe, Brazil; Formación Jiufotang (Aptiano), Grupo Jehol, de Liaoning occidental, China; y en las capas rojas (Cenomaniano) de la region de Kem Kem, Marruecos. Un cráneo incompleto hallado en la Formación Javelina (Maastrichtiano), Texas también muestra varios caracteres de los tapejáridos y podría per- tenecer a este clado. A pesar que la información es aún limitada la presente distribución de los Tapejaridae su- giere que este clado de pterosaurios no fue exclusivo del Gondwana y estuvo más ampliamente distribuido de lo que previamente se había conocido. Keywords. Pterosauria. Pterodactyloidea. Tapejaridae. Phylogenetic analysis. Cretaceous. Palabras clave. Pterosauria. Pterodactyloidea. Tapejaridae. Análisis filogenético. Cretácico. Introduction the same deposit (Tapejara wellnhoferi Kellner) was discovered and, together with Tupuxuara, classified In 1988 a toothless pterosaur from the Early in a new clade named Tapejaridae, whose relation- Cretaceous Santana Formation, Tupuxuara longi- ships with other pterodactyloid groups remained cristatus Kellner and Campos, was described and re- uncertain (Kellner, 1989). Based on other specimens garded to represent an (at that time) unknown clade from the Santana Formation, Kellner and Campos of the Pterodactyloidea (Kellner and Campos, 1988). (1992) suggested that tapejarids and azhdarchids One year later, a second toothless flying reptile from were in sister group relationship, a hypothesis that was supported by latter studies (Kellner and Hasegawa, 1993; Kellner and Langston, 1996; Kellner, 1996; 2001; 2003). The monophyly of the 1 Paleovertebrate Sector, Department of Geology and Paleonto- Tapejaridae was accepted by subsequent workers logy, Museu Nacional/UFRJ; Quinta da Boa Vista, São Cristóvão, Rio de Janeiro, RJ, 20940-040 Brasil. Fellow CNPq. (e.g., Unwin, 1995; Wellnhofer and Buffetaut, 1999; [email protected] Frey and Tischlinger, 2000; Wang and Zhou, 2002), ©Asociación Paleontológica Argentina AMGHB2-0002-7014/04$00.00+.50 522 A.W.A. Kellner but has been recently questioned (e.g.; Unwin and Lü, 1997; Unwin, 2000; 2003). Here I review the evidence pro and against the monophyly of the Tapejaridae. I also present new in- formation on several tapejarid and closely related species and present a new phylogenetic analysis that includes some recently described taxa tentatively re- ferred to this and related pterodactyloid clades. Abbreviations. BSP-Bayerische Staatssammlung fur Paläontologie und Historische Geologie, Munich. MCT-Museu de Ciências da Terra, Rio de Janeiro. MN-Museu Nacional/UFRJ, Rio de Janeiro. DNPM- National Department of Mineral Production, Rio de Janeiro. SMNK-Staatliches Museum für Naturkunde, Karlsruhe. Phylogenetic analysis In order to access the question of tapejarid mono- phyly, I performed a phylogenetic analysis which is an extension of the study published by Kellner (2003), with the addition of taxa and characters (see appendices). Among the taxa added are Thalassodro- meus sethi Kellner and Campos, regarded as a tape- jarid (Kellner and Campos, 2002a) and Zhejiangop- Figure 1. Strict consensus cladogram of the most parsimonious terus linhaiensis Cai and Wei, first regarded as a mem- trees recovered in the cladistic analysis. Outgroups were exclud- ber of the Nyctosauridae (Cai and Wei, 1994) but re- ed from the figure / Cladograma del consenso estricto de los tres ár- boles más parsimoniosos obtenidos en el análisis cladístico. Los grupos classified as an azhdarchid by Unwin and Lü (1997). externos fueron excluidos de la figura. 1, Pterodactyloidea; 2, Archae- The search was executed using PAUP 4.0b10 for Mi- opterodactyloidea; 3, Dsungaripteroidea; 4, Pteranodontoidea; 5, crosoft Windows (Swofford, 2000). The large dataset Tapejaroidea; 6, Dsungaripteridae; 7, Azhdarchoidea; 8, Tapeja- (3 outgroups + 42 pterosaur taxa; 80 characters, seve- ridae; 9, Azhdarchidae. See text for details. ral multistate) dictated the heuristic search option. Characters were given equal weight and treated un- ordered (ACCTRAN setting). The first search con- length of the skull, and rostrum downturned. With ducted by PAUP including all three outgroups the discovery of Tupuxuara leonardii Kellner and (Ornithosuchus longidens (Huxley), Herrerasaurus is- Campos and Tapejara imperator Campos and Kellner, chigualastensis Reig and Scleromochlus taylori Wood- it became clear that the downward projected rostrum ward) produced 6229 equally parsimonious trees is only prominent in Tapejara and is now regarded as with a length of 177 steps. When Scleromochlus taylori an apomorphy of that genus (character 4.1). The is excluded, PAUP finds 2095 trees (with 177 steps; species Sinopterus dongi Wang and Zhou, from Creta- consistency index=0.7627; retention index=0.9099; ceous strata of the Liaoning province in China and rescaled consistency index=0.6940). In both cases the assigned to the Tapejaridae also has a downturned topology of the consensus tree is the same (figure 1). rostral end, but shows a smaller inclination than Ta- The result indicates that the Tapejaridae are a mono- pejara (Wang and Zhou, 2002). phyletic entity that has a sister group relationship Compared to the length between the tip of the with the Azhdarchidae. premaxilla and the squamosal, the nasoantorbital fenestra in tapejarids is longer than in any other pte- rosaur (character 8.2), reaching about 45% in Tapejara Discussion of tapejarid monophyly wellnhoferi (figure 3) and Sinopterus dongi, half the size in Thalassodromeus sethi (figure 5) around 55% in When the Tapejaridae was erected, Kellner (1989) Tupuxuara leonardii (figure 4) and being largest in presented three features diagnosing this clade of Tapejara imperator (around 60%, figure 2). Although toothless pterosaurs: large sagittal crest on the ante- no detailed measurements are possible for Tupuxuara rior part of the skull extending backwards, very large longicristatus (Kellner and Campos, 1988) and the Ja- nasoantorbital fenestra occupying nearly half the velina tapejarid (figure 7, see section Composition of AMEGHINIANA 41 (4), 2004 Relationships of Tapejaridae (Pterosauria) 523 Figure 3. Skull and lower jaw of Tapejara wellnhoferi. Abbreviations: d, dentary; dsc, dentary sagittal crest; lpj, lacrimal process of the jugal; m, maxilla; naof, nasoantorbital fenestra; or, orbit; pm, premaxilla; pmsc, premaxillary sagittal crest. Scale bar = 50 mm / Cráneo y mandíbula de Tapejara wellnhoferi. Abreviaturas: d, dentario; dsc, cresta sagital del dentario; lpj, proceso lacrimal del yugal; m, maxilar;